grain size analyzers: results of an intercomparison study
TRANSCRIPT
HAL Id: hal-02454916https://hal.archives-ouvertes.fr/hal-02454916
Submitted on 6 Mar 2020
HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.
Copyright
Grain size analyzers: results of an intercomparison studyHugo Lepage, Matthieu Masson, Doriane Delanghe, Chloé Le Bescond
To cite this version:Hugo Lepage, Matthieu Masson, Doriane Delanghe, Chloé Le Bescond. Grain size analyzers: re-sults of an intercomparison study. SN Applied Sciences, Springer Verlag, 2019, 1 (9), pp.1100.�10.1007/s42452-019-1133-9�. �hal-02454916�
SN Applied SciencesGrain size analyzers: results of an intercomparison study
Manuscript Number: SNAS-D-19-01905R1
Full Title: Grain size analyzers: results of an intercomparison study
Article Type: Research Article
Section/Category: Earth and Environmental Sciences
Funding Information:
Abstract: In sciences involving soil and sediment, particle size distribution (PSD) has been one of the parameters given most attention over the past few decades. Formerly measured by sieving and sedimentation techniques, it is nowadays routinely characterized by the laser diffraction method (LDM). Many manufacturera develop particle size analyzers using LDM, but each device is characterized by specific parameters that can lead to different PSD. At the Rhône Sediment Observatory, suspended particulate matter collected along the Rhône River are analyzed for PSD by four different LDM devices. Analyses were conducted on certified materials and sediment samples for each device. The tests highlighted the difficulty of accurately characterizing PSD, even in the case of certified materials. First, differences observed for a specific device were linked to the heterogeneity observed in the subsamples due to the presence of organic materials such as tree leaves. Second, the difference regarding the certified materials was linked to the laser diffraction method which leads in some cases to underestimating clay content and sand. Third, the main difference observed between the devices was linked to sonication. The results demonstrate that its power is rarely investigated and that it has a considerable impact when used. However, despite significant differences, the trend was similar for each device, with accurate characterizations of the main modal class in most cases. Thus, in the absence of exact knowledge of parameters such as sonication power and pump speed, it is recommended to compare only the trends of the results obtained from the different devices.
Keywords: Particle size distribution; laser diffraction; ultrasound sonication; suspended particulate matter; Sediment
Corresponding Author: Hugo LepageInstitut de Radioprotection et de Sûreté NucléaireFontenay aux Roses, FRANCE
Corresponding Author Secondary Information:
Corresponding Author's Institution: Institut de Radioprotection et de Sûreté Nucléaire
Corresponding Author's Secondary Institution:
First Author: Hugo Lepage
First Author Secondary Information:
Order of Authors: Hugo Lepage
Matthieu Masson
Doriane Delanghe
Chloé Le Bescond
Order of Authors Secondary Information:
Author Comments:
Response to Reviewers: Dear Editor,
Please find enclosed the corrected manuscript according to the reviewer comments. A global reading of the paper was conducted and errors in the language were corrected.
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Kindly acknowledge the receipt of the same.
Your sincerely,
On behalf of the authors,Hugo LepageCorresponding authors,IRSN, PSE-ENV/SRTE/LRTACentre de Cadarache, bat 159, 13115 St Paul lez Durance, France [email protected]
Suggested Reviewers: Nele De BelieMagnel Laboratory for Concrete Research, Department of Structural Engineering, Faculty of Engineering and Architecture, Ghent University, Technologiepark Zwijnaarde 904, 9052 Ghent, Belgium [email protected] on the same field
Magdalena RyzakInstitute of Agrophysics, Polish Academy of Sciences, Lublin, Poland [email protected] on the same field
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Manuscript
Click here to view linked References
Click here toaccess/download;Manuscript;Manuscript_corrected.docx
1
12345678 9
1011121314151617181920 21 222324252627282930313233343536373839404142434445464748495051525354555657585960 61 62636465
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Grain size analyzers: results of an intercomparison study
Hugo Lepage3-*, Matthieu Massonb, Doriane Delanghec, Chloé Le Bescondb
a Institut de radioprotection et de Sûreté Nucléaire (IRSN), PSE-ENV, SRTE/LRTA, BP 3,13115 Saint-Paul-
lez-Durance, France
b Irstea, UR RiverLy, centre de Lyon-Villeurbanne, 5 rue de la Doua CS 20244, 69625 Villeurbanne, France
c Aix Marseille Univ, CNRS, IRD, INRA, Coll France, CEREGE, Aix-en-Provence, France
*Corresponding author: [email protected] - 0442199462 - httDs://orcid.org/0000-0002-2260-859X
ABSTRACT
In sciences involving soil and sediment, particle size distribution (PSD) has been one of the parameters
given most attention over the past few decades. Formerly measured by sieving and sedimentation
techniques, it is nowadays routinely characterized by the laser diffraction method (LDM). Many
manufacturers develop particle size analyzers using LDM, but each device is characterized by specific
parameters that can lead to different PSD. At the Rhône Sediment Observatory, suspended particulate
matter collected along the Rhône River are analyzed for PSD by four different LDM devices. Analyses
were conducted on certified materials and sediment samples for each device. The tests highlighted the
difficulty of accurately characterizing PSD, even in the case of certified materials. First, differences
observed for a specific device were linked to the heterogeneity observed in the subsamples due to the
presence of organic materials such as tree leaves. Second, the difference regarding the certified
materials was linked to the laser diffraction method which leads in some cases to underestimating clay
content and sand. Third, the main difference observed between the devices was linked to sonication.
The results demonstrate that its power is rarely investigated and that it has a considerable impact
when used. However, despite significant differences, the trend was similar for each device, with
accurate characterizations of the main modal class in most cases. Thus, in the absence of exact
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
2
knowledge of parameters such as sonication power and pump speed, it is recommended to compare
only the trends of the results obtained from the different devices.
Key word
Particle size distribution, laser diffraction, ultrasound sonication, suspended particulate matter,
sediment
Article Highlights
• Differences on the particle size distribution were always observed even with the certified
samples.
• The use of ultrasound sonication increases or triggers differences in the particle size
distribution.
• The lower is the proportion of a textural class (clay/silt/sand), the higher is the relative
difference.
1. INTRODUCTION
Much research has been conducted over recent decades to understand the dynamics and
behavior of sediment and suspended particles in rivers. Understanding the processes that affect the
transport of these particles is crucial for ecological, economic and societal purposes, as the world's
rivers are becoming more anthropized [1,2]. The investigation of their nature and origins has now
become essential [3]. Among the different parameters used to characterize particles when studying
their behavior (e.g., mineralogy, shape, color), particle size remains one of those most investigated
[4,5] and it is often an issue in geoscience studies.
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
3
The characterization of particle size distribution (PSD) can be performed using several
techniques such as sieving, by pipette based on Stokes diameters, and laser diffraction [6-8]. The
choice of the optimum grain-size analysis technique depends on the aim of the study and the type of
sediment [8]. For loamy sediments, laser diffraction instruments produced the best results for the
various criteria [8]. The main differences observed between the laser diffraction method (LDM) and
the other methods are related to particle morphology and clay mineralogy [5,9]. Clay content can be
underestimated in suspended sediments [10].
The LDM is widely used on sediment and suspended particles as it covers a wide range of grain
sizes and requires short analysis time [5,8,11]. For example, particle size analyses are processed on
cores recovered from marine, estuarian and coastal zones for paleoceanographic and sea level
reconstructions [12-15] and paleo-environmental questions associated with human history [16-18].
PSD is now commonly used to improve soil and sediment fingerprinting [19] and to investigate the
concentration of trace metals in sediments. Apart from the geosciences, many domains such as
forensic studies [20], pharmaceuticals and industry use this technology and require greater detail on
the data acquired from different lasers.
Nowadays, many devices from different manufacturers are used to measure PSD by LDM, with
different designs and software applications, leading to slightly different distributions [8,21].
Bieganowski et al. (2018) suggested that about 24 factors can influence the PSD results and lead to
uncertainty or error. Differences arise as all these parameters vary between devices and laboratories,
since resolving the issue of the homogeneous resuspension of sediment is a complex [22]. For example,
it has been demonstrated that the use of ultrasonic power can destroy aluminum and silica particles
due to the dissolution mechanism [23] and that the addition of oxidizing solutions to remove organic
matter can also alter clay particles such as vermiculite [24]. However, intercomparisons of PSD
measurements were performed more between the different PSD techniques described above than
between different LDM devices. Nevertheless, differences were observed for the PSD and the
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
4
proportion of clay measured by the devices of different manufacturers [8,10]. Unfortunately, in the
study conducted by Goossens (2008) on loamy sediments (sieved at 90pm), the impact of sand
particles and the use of ultrasound sonication (US) has not been addressed.
To investigate PSD, most of these studies used statistical parameters such as mean, median
and standard deviation to summarize PSD or Kurtosis and the Skewness index to describe the variability
of PSD [25,26]. However, such parameters might not be sufficient to describe PSD in a non-lognormal
or multimodal distribution [27,28]. Suspended sediment PSDs are mainly multimodal and the location
of modes can be related to the transportation processes and the origins of the particles [28]. Also,
modes can be used to investigate the dynamics of sediment deposition [29]. Several mathematical
solutions were developed to summarize and describe raw PSD such as two or five parameters Log-
normal equations [30,31], demodulation [32] and surface plots [28]. Unfortunately, these approaches
can be complex for people unfamiliar with advanced mathematics. This may explain why scientific
publications and technical reports are still published with common parameters [33-35]. Also,
depending on the study context, it is not necessary to rely on advanced solutions as common
parameters may be sufficient for investigating PSD.
Since 2009, within the context of the Rhône Sediment Observatory (OSR) program, scientists
have studied the transport of particles and their associated contaminants along the entire course of
the Rhône River (from Lake Geneva to the Mediterranean Sea), the main source of sediments in the
northern Mediterranean [36]. All these investigations are well-documented due to the presence of
several monitoring stations located along the Rhône River and in its main tributaries [37]. Particle
samples for chemical analysis are routinely collected. Analyses of PSD are systematically conducted
not only to study the transport of particles but also to improve the interpretation of contaminant
concentrations and they can also be used for transport modelling. For example, the use of different
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
5
sampling techniques for suspended particle sampling (particle trap and continuous flow
centrifugation) leads to a bias in particle size [38]. In the case of large environmental monitoring
networks such as the OSR program, PSD analyses are processed by an array of laboratory lasers.
It is therefore crucial to evaluate the potential biases induced by the use of different devices
in the PSD results. The Intercomparison was conducted by studying the differences observed on
common parameters, such as the proportion of clay/silt/sand resulting from the analysis - with and
without US - of reference standards and samples collected in the field. The objective was not to define
which device is the most accurate but quantify the relative difference that might be observed and how
the results can be interpreted and compared. Answering this question is crucial for the organizations
responsible for river surveys, as we provide tools for comparisons between all the PSDs analyzed within
the OSR.
2. MATERIALS AND METHOD
2.1. Particle size analyzers
In the OSR program, grain size is assessed using the following devices: a Beckman Coulter LS
13 320 (Beckman Coulter, Fullerton, CA, USA), a Cilas 1190L (Cilas Company Ltd., Orléans, France), a
Malvern Mastersizer 2000 (Malvern Instruments Ltd., Malvern, UK) and a Sequoia LISST-Portable|XR
(Sequoia Scientific, Bellevue, WA, USA). These devices will be referred to as Coulter, Cilas, Malvern and
Portable, respectively, for the rest of the article. As expected, the main characteristics of these devices
differ (Table 1).
2.2. Protocol of intercomparison
The intercomparison of the devices was conducted using certified reference materials (CRMs) and
natural samples from the Rhône River and its tributaries:
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
6
• Measurements were performed on CRMs to investigate the accuracy of the different devices
and highlight the differences between them. Three CRMs were selected in order to accurately
represent the ranges of suspended particle size, mostly silt-size [38], observed in the Rhône
River: BCR-066 (0.35-3.5 pm) and BCR-067 (2.4-32 pm) [39,40]; SRM1003c (25-50 pm) [41].
• Artificial multimodal distributions were created to assess the capacity of the device to
efficiently characterize the different modes. To do this, three mixtures of the 3 CRMs were
prepared by weighing masses in order to reach the following proportions (BCR-066/BCR-
067/SRM1003c in %): Mix_A: 25/50/25, Mix_B: 50/25/25 and Mix_C: 25/25/50.
• As CRMs are mostly composed with glass-quartz particles, they are not qualitatively
representative of the suspended particulate matter (SPM) encountered in the Rhône River
[42], especially regarding their aggregation properties. Therefore, an intercomparison was also
conducted on the SPM samples collected in the observatory. Four samples were collected
using different methods at different stations to characterize the diversity observed in this
observatory:
o Arc River (45°33'44.6"N 6°12'22.8"E): lag deposits that settled during a flood event were
collected manually with a plastic spatula. This sediment was sieved in order to keep only a
sandy fraction between 100 pm and 2 mm.
o Azergues River (45°56'11.5"N 4°43'24.6"E): bottom sediment collected by a sediment
dredge (Eckman),
o Jons monitoring station in the middle Rhône River (45°48'42.3"N 5°05'09.2"E): SPM
collected by particle trap (Masson et al., 2018) during baseflow in March 2017,
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
7
o Barcarin station in the downstream Rhône River (43°25'12.5"N 4°44'50.7"E): SPM collected
using a Niskin bottle 1m above the surface during a flood event in November 2017. This
was the only liquid sample comprising SPM in Rhône River water.
• The last step was conducted on a mixture of the samples collected in the Azergues and Arc
Rivers to control the proportion of modes and the impact of coarse particles. The mixtures
were prepared by weighing dry particles in the following proportions (Azergues/Arc in %):
Mix_1: 95/5, Mix_2: 90/10, Mix_3: 85/15, Mix_4: 80/20 and Mix_5: 70/30.
Each sample was homogenized then split into 4 subsamples before being sent to the different
laboratories for PSD analysis. For each sample, measurements were repeated on 2 or 3 distinct aliquots
depending on the quantity available. Prior to measurement, the subsamples were dispersed in water
according to Arvaniti et al. (2014).
The Mie model was used when possible as the optimal model, as it is more useful in samples
characterized by fine silt and clay content [43]. The Coulter device was the only device in this
intercomparison that uses both models. Depending on the device, the refraction indexes were chosen
to be similar and close to silica as the variation of this index can result in different PSDs [11]. The other
parameters of the analysis method (Table 1), such as stirrer time and speed, pump speed adapted to
the volume of the bowl, ultrasonic duration, and measurement duration, were determined by the
users to obtain the best possible results (these parameters were optimized for each device with
particulate samples collected in the OSR program).
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
8
After a series of tests on the certified standards (data not presented), measurements were
performed by conforming to the best obscuration range given by the manufacturer (Table 1). The
results of the tests show differences occurring outside those ranges.
Each subsample was analyzed at least three times without US then three times with US to
investigate the repeatability of the measurements and the effect of sonication on the natural samples.
2.3. Comparison parameters
One of the main difficulties in conducting an intercomparison on PSD is the need to use
synthetic parameters, as the raw distributions cannot be directly compared due to channel
inhomogeneity. In the following study, intercomparison will be conducted on several of the
parameters used most frequently in environmental studies and especially in the OSR program:
• The median (d50) that synthesizes the distribution by one factor. However, this parameter is
not the most relevant in the case of multimodal distribution, as is mostly encountered in rivers
[28,31].
• The proportions of clay (<1.95pm), silt (1.95 - 62.5pm) and sand (62.5 - 2000pm) are
commonly used for normalization, as contaminants are mostly fixed on clay and silt. Again, this
does not give information on distribution shape, especially in the case of multimodality [28]
and the range of the classes can vary depending on the scale used [26,44]. This separation into
3 classes will be referred to as "simple classes" for the rest of the text.
• The proportions of detailed classes follow the Doeglas scale [45]: clay (<1.95pm), very fine silt
(V_F_Silt: 1.95 - 3.91pm), fine silt (F_Silt: 3.91 - 7.81pm), medium silt (M_Silt: 7.81 -
15.63pm), coarse silt (C_Silt: 15.63 - 31.25pm), very coarse silt (V_C_Silt: 31.25 - 62.5pm),
very fine sand (V_F_Sand: 62.5 - 125pm), fine sand (F_Sand: 125 - 250pm), medium sand
(M_Sand: 250 - 500pm), coarse sand (C_Sand: 500 - 1000pm), very coarse sand (V_C_Sand:
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
9
1000 - 2000pm). It allows a good and comparable description of the distribution but the higher
classes are less-well adapted for statistical analyses than previous parameters. This separation
will be referred to as "detailed classes" for the rest of the text.
2.4. Spécifie case of the LISST Portable|XR device
Due to its lower range of PSD than the other devices (Table 1), and as it cannot characterize
the particles with a diameter lower than 0.34 pm, this LDM device is sensitive to rising tails [46-48].
Rising tails can be linked to the refractive index for small particles [49] or shape effects [12], and most
studies recommend removing the lower classes and keeping the larger size classes that are consistent
[46]. Thus, as rising tails were observable for almost all the samples, the first 5 classes (< 0.72pm) of
all the samples measured by the Portable were removed.
However, for this study, both raw and corrected PSDs were used to investigate the impact of
rising tails on the PSD. Corrected PSD will be referred to as "Portable_C" for the rest of the text. Only
the results of Portable_C will be presented in the graph for the sake of clarity and the results without
correction are available in the Electronic Supplementary Material. Also, the presence of bubbles during
the analysis led to the presence of modes in the higher classes. This was especially relevant for the
CRMs that should not be characterized by such coarse particles. Corrections were performed on
references (BCR-066 & BCR-067) and a mixture of references (Mix_C), by deleting the coarser classes.
3. RESULTS
3.1. Accuracy of the devices
The first part entailed measuring the CRMs. For this step, the use of US had a negligible impact
as CRMs are characterized by non-cohesive materials (quartz). Only the results with US will be
presented in this chapter.
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
10
First, the results obtained by each device from the different aliquots of the CRMs (Fig. 1) were
characterized by low standard deviations (SD) except for mixture C (Mix_C) measured by the Coulter
(high SD on clay and coarse silt). For this device with this mixture, the mean d50 ranged from 2.4 ± 0.1
^m to 6.2 ± 1.4 ^m for the three aliquots measured and the mean proportion of clay ranged from 27
± 2 % to 42 ± 1 % (Electronic Supplementary Material).
Second, the values measured by the devices were compared to the certified values (Table 2 -
Fig. 1). For the finest CRM (BCR-066), all the devices overestimated the d50 (Table 2) and
underestimated the proportion of clay (Fig. 1). The absolute difference ranged from 13 % to 34% for
the clay proportion, which represented 15% to 40% of the relative difference. For BCR-067 and
SRM1003c, the measured d50s were similar to the certified parameters with a relative difference lower
than 20% (Table 2). For the simple classes, the relative difference was also lower than 20% but each
device measured particles in classes in which the CRMs should not be observed. Indeed, all the PSDs
measured by the devices were more spread out with a low amount (<10%) of particles observed in the
finer and coarser classes than the expected classes of the CRMs. This difference was clearly observed
on the detailed classes (Fig. 1) and the relative difference exceeded 20% for all the devices on the
classes certified by the BCR-067 (from 25% to 175%). For the SRM1003c, the difference exceeded 10%
only once with the proportion of very coarse silt measured by the Portable_C device (38%) (Fig. 1).
Thus, all the devices underestimated the proportion of particle in the modal class (Fig. 1).
The intercomparison of the devices was conducted on those results. The d50s measured on
the CRMs by the four devices were similar whereas significant differences were observed between the
devices for the mixtures (Table 2). For instance, the d50 measured by the Cilas for mixture C (11.4 ±
0.5 ^m) was more than twice as high as the d50 measured by the Coulter (4.6 ± 2.3 ^m). Modal classes
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
11
of the CRMs (unimodal) and the mixtures (bimodal) were accurately characterized by all the devices
with the exception of mixture B measured by the Coulter. For this mixture, the PSD measured by the
Coulter was unimodal and no particles coarser than medium silt was measured whereas it represented
more than 20% for the other device (Fig. 1). Thus, the proportion of clay measured by the Coulter was
significantly higher than the other devices for this mixture (54% vs 30% to 33% - Fig. 1). For the other
samples, the proportions measured by the devices were similar (<10% absolute difference) except for
the finest CRM (BCR-066). The proportions of clay measured by the Cilas and the Coulter on this CRM
(74% and 70%, respectively) were significantly higher than the two other devices (53% and 58% - Fig.
1).
3.2. Natural samples
The first natural sample, collected directly in the water flow with a Niskin bottle at Barcarin,
was mostly silt sized with a non-negligible proportion of clay (Fig. 2 - Table 3). The PSD measured by
the different devices were similar, so as for the d50 (Table 4). A difference was observed only for the
Coulter which measured a significant and higher proportion of clay (+11-13%), as observed previously
for Mixtures B and C (Fig. 1). The d50 measured by the Coulter was strictly lower than the other devices
(Table 4). Sonication was used only on the Cilas and had a negligible impact on this sample (Electronic
Supplementary Material).
The sample collected by particle trap at Jons was coarser than the SPM from Barcarin (Fig. 2),
with a significant proportion of sand (up to 27 % - Table 3). The differences between the devices
observed on the measures performed without US remained lower than 10% (Fig. 2) and the d50s were
similar (Table 4). As observed previously, the d50 and the clay content measured by the Coulter were
respectively lower and higher than the results obtained by the two other devices. The use of US led to
significant change in the PSD for the Portable_C (from 20% to 3% of very fine sand), while PSD
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
12
remained similar for the Coulter and Cilas. Thus, the proportion of silt was significantly higher (+20%)
for the Portable_C with sonication than for the other devices (Table 3).
The bottom sediment collected in the Azergues River was characterized by a mix of silt and
sand particles (Table 3), with the main mode observed between the coarse silt and the very fine sand
(Fig. 2). In absence of US, the trend was similar for each device and only the Portable_C was
characterized by a higher proportion in the silted classes and a lower proportion of sand. The
proportion of silt measured by Portable_C (72.6 ± 9.9 %) was significantly higher than the other devices
(from 44.2 to 51.3 %) but characterized by a high standard deviation. The use of US on this sample led
to a significant variation for all the devices and to an increase of silt particles (+ 6% to 37%) for all the
devices (Table 3). However, this variation was greater for the Cilas (Fig. 2) (+37% of silt) than for the
three other devices (+6-16%). The proportions of the classes after US were different and separated
into two groups: Cilas and Portable_C with almost 90% silt, with coarse silt as the modal class, and
Coulter and Malvern characterized by a large proportion of sand (35-43%) and very coarse silt as the
modal class. Once again, the differences observed for the d50 were significant whether US was used
or not (Table 4). For instance, d50s measured by the Malvern with and without US (respectively 40.0 ±
5.4 pm and 71.6 ± 5.8 pm) were two times higher than on Portable_C (20.9 ± 1.8 pm and 37.9 ± 9.1
pm).
The last sample, the lag deposit collected in the Arc River, was characterized by fine sands (Fig.
2). For the four devices, the proportion of sand was higher than 80% and reached 100% for the Malvern
(Table 3). The Coulter and Portable_C distributions were characterized by higher uncertainty intervals.
These devices also measured a higher mean proportion of silt particles compared to the other two
devices, resulting in a significantly lower d50 (Table 4). However, despite this difference, the trend was
similar between the devices with the maximum proportion observed in the fine sand class (Fig. 2). The
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
13
PSDs measured by the Malvern were less spread than for the other device while the proportion of sand
classes measured by the Cilas was more spread with a lower proportion in the modal class (F_Sand)
and a higher proportion in the coarser sand class (M_Sand). Finally, as this sample was mostly sandy,
US had a negligible impact on the distribution, as observed for the d50 (Table 4).
3.3. Mixtures of natural samples
Measures were then performed on the mixtures of the Arc and Azergues samples. As expected,
the increase of the proportion of the coarser sample (from mixture 1 to 5) led to a higher d50 (Fig. 3)
and higher proportions in the coarser classes (Electronic Supplementary Material) whatever sonication
was used or not. However, breaks were observed for the Coulter (Mix_1), the Malvern (Mix_4 and
Mix_5) and the Portable_C (Mix_3 and Mix__5).
The differences observed between the devices were similar to the difference observed for the
Azergues sample as it represented more than 70% of the mixture. Thus, the parameters were different
both with and without US (Fig. 4). Significant differences were observable for d50 and with the
proportion of silt and sand, whereas the proportion of clay remained lower than 5% for each device
(Fig. 4). The position of the main mode was also different, both without (simple and detailed classes)
and with US (detailed classes). Using US led to the strong variation for simple classes measured by the
Cilas and the Portable_C (+35% silt) whereas it had no effect for the Coulter (Fig. 3-4). US also affected
the Malvern device (+20% silt). The trends and the differences between the devices were similar
whatever the proportion of sand in the samples.
4. DISCUSSION
The measures performed by the 4 devices on the different samples (CRMs, sediments and
associated mixtures) allowed determining their respective capacities in terms of measurement quality.
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
14
The results demonstrated that several différences were observed despite the good and reproducible
characterization of the main/modal classes and their distribution by the devices. These differences can
be separated into three groups: intra-variation, variation with the certified materials, and variation
between the devices.
4.1. Intra-variation
The intra-variation is the difference observed for the aliquots of the same subsample measured
by the same device. For the CRMs, this difference was observed only in mixture C measured by the
Coulter (Fig. 1) and is linked to the operator subsampling operation. Subsampling appears to be highly
operator-dependent and can infer variations on the results of the same sample that are even higher
than the analytical differences observed [5]. The intra-variation of the natural samples was observed
mostly for coarse classes (very coarse silt to fine sand) measured by the Coulter and the Portable_C
devices (Fig. 2). This was observed both without and with US and was mainly related to the
heterogeneity of the sample particles. Indeed, during the subsampling some samples contained coarse
organic particles such as tree leaves that modified the proportions between modes. For the
Portable_C, it was also related to the presence of bubbles as explained in the chapter 2.4. To improve
the characterization of sand particles, the pump speed of this device was increased for coarse samples
and triggered the formation of bubbles. It is therefore relevant to perform measures on multiple sub-
samplings of a sample (at least three) to characterize a more precise mean PSD and to be careful with
the presence of coarse organic materials and bubbles. Also, the sieving of coarse particles could be a
solution for decreasing the pump speed.
4.2. Variation with the CRMs
The measures performed on the CRMs for all the devices presented differences with the
certified values. First, they underestimated the proportion of clay in the finest reference (BCR-066) and
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
15
the different associated mixtures (Fig. 1). This underestimation was linked to the particle morphology
and the clay mineralogy [9] affecting the PSD measured by LDM. Similar observations had already
performed by [10,21]. Second, the PSDs measured by all the devices were more spread out than the
associated references or mixtures. Also, particles were observed in classes where they should not have
been present. For example, all the devices measured clay particles with the BCR-067 and SRM1003c
(Fig. 1). However, the proportions in these classes were lower than 10%. This difference might be
related to the presence of particles not washed during cleaning. Despite the measure of the
background PSD with tap or dionized water, some particles might have been trapped in the system
and resuspended during other measurements. This difference might also be linked to the different
methods used to characterize the particle size for their certification. BCR-066 and BCR-067 were
characterized by the Pipette method [39] which might have triggered differences with the LDM, as
observed with the proportion of clay.
To summarize, all 4 devices exhibit the same trends in accuracy from small to coarser particles.
However, deviations on the d50 remain roughly lower than 10% for the standard ranging from 4pm to
35pm. As explained previously, the devices do not accurately characterize clay, whatever the
proportion observed, with the exception of the total absence of clay. Higher relative differences are
observed when the proportion of clay is low (approx. 10%). For the silt, a small difference is observed
in the CRMs characterized by almost only silt particles. Below 30% silt, the difference can be significant
(from 20% up to 300%), especially for CRMs mainly characterized by clay and silt (BCR-066). Finally,
the absence of particles in any simple class is accurately characterized by the devices.
4.3. Intercomparison of the devices
The variation between the devices was observed on almost all the samples including the CRMs
and their mixtures and was directly related to the specificities of the devices. As explained previously,
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
16
each device is manufactured differently (water volume, laser cells, etc.). One of the main source of
différences is linked to the use of US and its power. Strong (Cilas, Portable_C) and weak (Coulter)
variations were therefore observed after using US for certain samples. The absence of variation for the
Coulter demonstrated that the US power might have been insufficient to separate the aggregates.
However, it is difficult to investigate US power on CRMs as they are mostly composed of non-cohesive
materials and, unfortunately, this parameter was not expressed in Joules by mL [50], so the power
used by the different devices was different (Table 1). This observation demonstrates the need to
describe the power and time of US used in future studies to ensure reproducibility, as expressed by
Bieganowski et al. (2018). Sonication is used to estimate the aggregation index of soil and sediment
samples [51] and the results obtained may not be comparable in the absence of such description.
Another strong difference was observed when the main mode was located close to an interval
between two classes, a phenomenon also observed for the CRM SRM1003c. The proportion of these
two classes could be significantly different between the devices which distribute the mode between
them (Fig. 1). Moreover, this distribution can be problematic when the interval is located between clay
and silt or silt and sand such as the Azergues sample (Fig. 2 - Table 3). Vigilance is crucial when
comparing such results.
Results on raw PSD and corrected PSD of the Portable were compared to investigate the impact
of the rising tails observed for the finest classes (Electronic Supplementary Material). The results
demonstrated that rising tails had a negligible impact on the proportion of the classes. Highest
difference was observed on the clay content of the Barcarin sample with 18% on the raw PSD and 11%
on the corrected PSD. Moreover, the potential characterization of a lower proportion of clay by
deleting the finest classes remained non-significant and the values were mainly in the range of the
values of the other devices. For this specific study, it was not relevant to remove the finest classes.
However, correction remained necessary for the coarse modes induced by the presence of bubbles
caused by the pump speed as explained above.
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
17
To evaluate the différences and allow users to determine those they can obtain with their
devices, figures 5, 6 and 7 summarize the observations by grouping all the measures performed on all
the natural samples and mixtures for all the parameters studied. Relative differences were then
estimated for each device.
For d50 (Fig. 5), the larger differences (up to 80% of relative difference with the mean d50)
were observed when the sample was a mix of silt and sand particles with a d50 ranging from 25 pm to
50 pm or from 50 pm to 80pm, with and without US, respectively. For samples with a high d50 (higher
than 150pm), the difference remained lower than 20% while it could exceed 20% for samples with a
low d50.
For the simple classes (Fig. 6), the relative difference was significant (>20%) when the
proportion was lower than 70% and 80% with and without US, respectively, and this difference
increased when the proportion decreased and was lower than 25%.
Similar observations were performed for the detailed class (Fig. 7), the relative difference
decreased as the proportion increased, but remained mostly higher than 20% whatever the proportion
and the US. Therefore, since multimodality was characterized in most cases, detailed classes should be
used to describe the trend of the PSD rather than comparing results from different devices on a
qualitative basis.
Finally, the Specific Surface Area (SSA) was estimated [52] based on the PSDs of the different
references and samples. However, as the proportion of clay was different between the devices, and
even non-significant, the SSAs estimated by them were all substantially different from each other. It
appeared that this parameter is difficult to reproduce on different devices and cannot be compared
directly.
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
18
5. CONCLUSION
The tests performed during this study demonstrated that the comparison of particle size
measured using different devices can be complicated since differences were observed on certified
materials and natural samples. Despite using similar protocols and samples, the results obtained for
the selected parameters (d50, simple and detailed classes) were characterized by differences that
could be significant. Those differences were directly related to the specificities of the devices, as each
device is manufactured differently (water volume, number of laser cells, etc.). In the absence of
information on measurement parameters such as US power in J.ml-1 and pump speed, the comparison
of PSDs from different studies can be erroneous, even for a similar device. It is therefore better to only
compare the trend (such as the modal classes or the presence of these types of particle) rather than
the measured proportion. Quantitative comparisons should only be performed if the parameters of
the different analyses are known.
Differences in d50 were observed even with CRMs. Deviation from the expected d50 observed
on all the lasers presented similar trends, with increases in particle size. Although there were
differences in the amplitude of uncertainty, they all estimated the d50 for particle-sizes above 10pm
within 10% uncertainty. However, for samples mostly characterized by clay and fine silt particles, the
d50 was overestimated. This is important, especially for natural samples where these fractions are
often present. Indeed, this information is required in numerous studies. Moreover, on natural
sediments mixed with clay silts and sands, the problem of sand is not trivial. Sand is often not seen in
these cases when clay size particles are present in quantity. Sieving sediments to remove the sand
might be a solution to eliminate the impact of coarse particles.
For the different classes, the main problem observed concerned the position of the modes.
Displaying the distribution of the classes greatly depends on the classification chosen relative to the
sediment grain size distribution. It is therefore crucial to define the localization of the modes before
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
19
conducting a comparison with results from different devices. Significant différences can be observed
in samples characterized by a PSD with modes between two classes (clay/silt, silt/sand or detailed
classes).
ACKNOWLEDGEMENTS
We gratefully acknowledge the following colleagues: Maxime Darbot, Marina Launay, Marie
Courtel, Stéphanie Gairoard and Brice Mourier for sample preparation and PSD analysis.
Funding
This study was supported by the Rhône Sediment Observatory (OSR), a multi-partner research
program partly funded by the Plan Rhône, and by the European Regional Development Fund (ERDF)
allocated by the European Union.
Compliance with ethical standards
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
REFERENCE
[1] Walling DE, Fang D. Recent trends in the suspended sediment loads of the world's rivers. Glob
Planet Change 2003;39:111-26. doi:10.1016/S0921-8181(03)00020-1.
[2] Frings RM, Ten Brinke WBM. Ten reasons to set up sediment budgets for river management.
Int J River Basin Manag 2018;16:35-40. doi:10.1080/15715124.2017.1345916.
[3] SedNet. Moving Sediment Management Forward. The Four SedNet Messages. 2014:16 pp.
https://sednet.org/download/Moving-Sediment-Management-Forward.pdf (accessed January
22, 2019).
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
20
[4] Wills BA, Napier-Munn T. Wills' Minerai Processing Technology. Wills' Miner Process Technol
2005:267-352. doi:10.1016/B978-075064450-1/50014-X.
[5] Bieganowski A, Ryzak M, Sochan A, Barna G, Hernadi H, Beczek M, et al. Laser Diffractometry in
the Measurements of Soil and Sediment Particle Size Distribution. Adv Agron 2018;151:215-79.
doi:10.1016/bs.agron.2018.04.003.
[6] Syvitski JPM. Principles, Methods and Application of Particle Size Analysis. Cambridge University
Press; 1991.
[7] Merkus HG. Particle Size Measurements - Fundamentals, Practice , Quality. Springer S. Springer
Netherlands; 2009. doi:10.1007/978-1-4020-9015-8.
[8] GOOSSENS D. Techniques to measure grain-size distributions of loamy sediments: a
comparative study of ten instruments for wet analysis. Sedimentology
2007;55:070921101149001-??? doi:10.1111/j.1365-3091.2007.00893.x.
[9] Di Stefano C, Ferro V, Mirabile S. Comparison between grain-size analyses using laser diffraction
and sedimentation methods. Biosyst Eng 2010;106:205-15.
doi:10.1016/j.biosystemseng.2010.03.013.
[10] Loizeau J -L, Arbouille D, Santiago S, Vernet J -P. Evaluation of a wide range laser diffraction
grain size analyser for use with sediments. Sedimentology 1994;41:353-61.
doi:10.1111/j.1365-3091.1994.tb01410.x.
[11] Arvaniti EC, Juenger MCG, Bernal S a., Duchesne J, Courard L, Leroy S, et al. Determination of
particle size, surface area, and shape of supplementary cementitious materials by different
techniques. Mater Struct 2014;48:3687-701. doi:10.1617/s11527-014-0431-3.
[12] Agrawal YC, Whitmire A, Mikkelsen OA, Pottsmith HC. Light scattering by random shaped
particles and consequences on measuring suspended sediments by laser diffraction. J Geophys
Res Ocean 2008;113:1-11. doi:10.1029/2007JC004403.
[13] Allen JRL, Haslett SK. Granulometric characterization and evaluation of annually banded mid-
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
21
Holocene estuarine silts, Welsh Severn Estuary (UK): coastal change, sea level and climate. Quat
Sci Rev 2006;25:1418-46. doi:10.1016/j.quascirev.2005.12.009.
[14] Bassetti MA, Berné S, Jouet G, Taviani M, Dennielou B, Flores JA, et al. The 100-ka and rapid sea
level changes recorded by prograding shelf sand bodies in the Gulf of Lions (western
Mediterranean Sea). Geochemistry, Geophys Geosystems 2008;9. doi:10.1029/2007GC001854.
[15] McCave IN, Thornalley DJR, Hall IR. Relation of sortable silt grain-size to deep-sea current
speeds: Calibration of the "Mud Current Meter." Deep Res Part I Oceanogr Res Pap 2017; 127:1-
12. doi:10.1016/j.dsr.2017.07.003.
[16] Ghilardi M, Psomiadis D, Andrieu-Ponel V, Colleu M, Sotiropoulos P, Longo F, et al. First
evidence of a lake at Ancient Phaistos (Messara Plain, South-Central Crete, Greece):
Reconstructing paleoenvironments and differentiating the roles of human land-use and
paleoclimate from Minoan to Roman times. Holocene 2018;28:1225-44.
doi:10.1177/0959683618771473.
[17] Giaime M, Avnaim-Katav S, Morhange C, Marriner N, Rostek F, Porotov A V., et al. Evolution of
Taman Peninsula's ancient Bosphorus channels, south-west Russia: Deltaic progradation and
Greek colonisation. J Archaeol Sci Reports 2016;5:327-35. doi:10.1016/j.jasrep.2015.11.026.
[18] Ghilardi M, Psomiadis D, Cordier S, Delanghe-Sabatier D, Demory F, Hamidi F, et al. The impact
of rapid early- to mid-Holocene palaeoenvironmental changes on Neolithic settlement at Nea
Nikomideia, Thessaloniki Plain, Greece. Quat Int 2012;266:47-61.
doi:10.1016/j.quaint.2010.12.016.
[19] Laceby JP, Evrard O, Smith HG, Blake WH, Olley JM, Minella JPG, et al. The challenges and
opportunities of addressing particle size effects in sediment source fingerprinting: A review.
Earth-Science Rev 2017;169:85-103. doi:10.1016/j.earscirev.2017.04.009.
[20] Pye K, Blott SJ. Particle size analysis of sediments, soils and related particulate materials for
forensic purposes using laser granulometry. Forensic Sci Int 2004;144:19-27.
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
22
doi:10.1016/j.forsciint.2004.02.028.
[21] Konert M, Vandenberghe J. Comparison of laser grain size analysis with pipette and sieve
analysis: a solution fo rthe underestimation of the clay fraction. Sedimentology 1997,'44:523-
35.
[22] Jonasz M, Fournier GR. Light Scattering by Particles in Water: Theoretical and Experimental
Foundations. Elsevier/Academic Press; 2007. doi:10.1016/B978-0-12-388751-1.X5000-5.
[23] Lu Y, Riyanto N, Weavers LK. Sonolysis of synthetic sediment particles: Particle characteristics
affecting particle dissolution and size reduction. Ultrason Sonochem 2002;9:181-8.
doi:10.1016/S1350-4177(02)00076-7.
[24] Mikutta R, Kleber M, Kaiser K, Jahn R. Review : Organic Matter Removal from Soils using
Hydrogen Peroxide , Sodium Hypochlorite, and Disodium Peroxodisulfate. Soil Sci Soc Am J
2005,'69:120. doi:10.2136/sssaj2005.0120.
[25] Folk RL. a Review of Grain-Size Parameters. Sedimentology 1966;6:73-93. doi:10.1111/j.1365-
3091.1966.tb01572.x.
[26] Blott SJ, Pye K. Particle size scales and classification of sediment types based on particle size
distributions: Review and recommended procedures. Sedimentology 2012;59:2071-96.
doi:10.1111/j.1365-3091.2012.01335.x.
[27] Roberson S, Weltje GJ. Inter-instrument comparison of particle-size analysers. Sedimentology
2014;61:1157-74. doi:10.1111/sed.12093.
[28] Beierle BD, Lamoureux SF, Cockburn JMH, Spooner I. A new method for visualizing sediment
particle size distributions. J Paleolimnol 2002;27:279-83. doi:10.1023/A:1014209120642.
[29] Visher GS. Grain Size Distributions and Depositional Processes. J Sediment Petrol
1969;39:1074-106. doi:10.1306/74D71D9D-2B21-11D7-8648000102C1865D.
[30] Gardner WR. Representation of Soil Aggregate-Size Distribution by a Logarithmic-Normal
Distribution. Soil Sci 1956:12-4. doi:10.1002/hep.23592.
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
23
[31] Fredlund MD, Fredlund DG, Wilson GW. An équation to represent grain-size distribution. Can
Geotech J 2000;37:817-27. doi:10.1139/t02-080.
[32] Lee BJ, Fettweis M, Toorman E, Molz FJ. Multimodality of a particle size distribution of cohesive
suspended particulate matters in a coastal zone. J Geophys Res Ocean 2012;117.
doi:10.1029/2011JC007552.
[33] Sadeghi SH, Singh VP, Kiani-Harchegani M, Asadi H. Analysis of sediment rating loops and
particle size distributions to characterize sediment source at mid-sized plot scale. Catena
2018;167:221-7. doi:10.1016/j.catena.2018.05.002.
[34] Ding W, Huang C. Effects of soil surface roughness on interrill erosion processes and sediment
particle size distribution. Geomorphology 2017;295:801-10.
doi:10.1016/j.geomorph.2017.08.033.
[35] Zhang Y, Sangster JL, Gauza L, Bartelt-Hunt SL. Impact of sediment particle size on
biotransformation of 17P-estradiol and 17P-trenbolone. Sci Total Environ 2016;572:207-15.
doi:10.1016/j.scitotenv.2016.08.004.
[36] Sadaoui M, Ludwig W, Bourrin F, Raimbault P. Controls , budgets and variability of riverine
sediment fluxes to the Gulf of Lions (NW Mediterranean Sea). J Hydrol 2016;540:1002-15.
doi:10.1016/j.jhydrol.2016.07.012.
[37] Poulier G, Launay M, Le Bescond C, Thollet F, Coquery M. Combining flux monitoring and
estimation to establish annual budgets of suspended particulate matter and associated
pollutants in the Rhône River from Lake Geneva to the Mediterranean Sea. Sci Total Environ
2019;658:457-73. doi:10.1016/J.SCITOTENV.2018.12.075.
[38] Masson M, Angot H, Le Bescond C, Launay M, Dabrin A, Miège C, et al. Sampling of suspended
particulate matter using particle traps in the Rhône River: Relevance and representativeness
for the monitoring of contaminants. Sci Total Environ 2018;637-638:538-49.
doi:10.1016/j.scitotenv.2018.04.343.
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
24
[39] Institute for Reference Materials and Measurements. Certified reference materials BCR-066.
2007.
[40] Institute for Reference Materials and Measurements. Certified Reference Material BCR-067.
2007.
[41] Kaiser D, Watters RL. Standard Reference Material® 1003c. Natl Inst Stand Technol 2011.
[42] Slomberg DL, Ollivier P, Radakovitch O, Baran N, Sani-Kast N, Miche H, et al. Characterisation of
suspended particulate matter in the Rhone River: Insights into analogue selection. Environ
Chem 2016;13:804-15. doi:10.1071/EN15065.
[43] Santiago S, Thomas RL, McCarthy L, Loizeau JL, Larbaigt G, Corvi C, et al. Particle Size
Characteristics of Suspended and Bed Sediments in the Rhone River. Hydrol Process
1992;6:227-40. doi:10.1002/hyp.3360060210.
[44] USDA. Soil Survey Manual. Handbook18. Washington, D.C.: Government Printing Office; 2017.
doi:10.1097/00010694-195112000-00022.
[45] Doeglas DJ. Grain-Size Indices, Classification and Environment. Sedimentology 1968;10:83-100.
doi:10.1111/j.1365-3091.1968.tb01101.x.
[46] Xi H, Larouche P, Tang S, Michel C. Characterization and variability of particle size distributions
in Hudson Bay, Canada. J Geophys Res Ocean 2014;119:3392-406. doi:10.1002/2013JC009542.
[47] Mikkelsen OA, Pejrup M. The use of a LISST-100 laser particle sizer for in-situ estimates of floc
size, density and settling velocity. Geo-Marine Lett 2001;20:187-95.
doi:10.1007/s003670100064.
[48] Many G, Bourrin F, Durrieu de Madron X, Pairaud I, Gangloff A, Doxaran D, et al. Particle
assemblage characterization in the Rhone River ROFI. J Mar Syst 2016;157:39-51.
doi:10.1016/j.jmarsys.2015.12.010.
[49] Andrews S, Nover D, Schladow SG. Using laser diffraction data to obtain accurate particle size
distributions: The role of particle composition. Limnol Oceanogr Methods 2010;8:507-26.
574
575
576
577
578
579
580
581
582
583
584
25
doi:10.4319/lom.2010.8.507.
[50] Yang XM, Drury CF, Reynolds WD, MacTavish DC. Use of sonication to determine the size
distributions of soil particles and organic matter. Can J Soil Sci 2009;89:413-9.
doi:10.4141/cjss08063.
[51] Jouon A, Ouillon S, Douillet P, Lefebvre JP, Fernandez JM, Mari X, et al. Spatio-temporal
variability in suspended particulate matter concentration and the role of aggregation on size
distribution in a coral reef lagoon. Mar Geol 2008;256:36-48.
doi:10.1016/j.margeo.2008.09.008.
[52] Santamarina JC, Klein K a, Wang YH, Prencke E. Specific surface: determination and relevance.
Can Geotech J 2002;39:233-41. doi:10.1139/t01-077.
Tablel Click here to access/download;Table;Table1.docx ±
Table 1: Main parameters of the grain size analyzers used within the OSR. n.a. not available.
Main characteristics of the devicesBrand Cilas Beckman Coulter Malvern SequoiaModel 1190L LS 13 320 Mastersizer 2000 LISST-Portable|XRRange of measured size (|am) 0.04 - 2500 0.04 - 2000 0.02 - 2000 0.34 - 500Number of channels 100 132 100 33Max. volume of the vessel (mL) 450 1000 1000 117
Main parameters set for analysisUltrasonic use 30s before during analysis during analysis 30s before analysis
and during analysisType of liquid and dispersant deionized water tap water tap water deionized water
without dispersant without dispersant without dispersant without dispersantDuration of a single measurement 60 s 60 s 10 s 60 sOptical model Mie Mie+Fraunhofer* Mie MieRefractive index 1,55 1,5 1,57 1,54Absorption coefficient 0,1 0,0 0,1 0,1Ultrasound sonication power 50% 6 on scale
of 8 levels100% 30%
Stirrer speed 150 RPM n.a. 1000 RPM n.a.Pump speed 120 RPM 80-90% 2500 RPM 20-25%Obscuration (Obs) or Optical Transmission (OT)
Obs: 5 - 25% Obs: 8-16% Obs: 5 - 25% OT: 0.75 - 0.95
* Fraunhofer used when particles are coarse (LS-13320 Manual)
Table2 Click here to access/download;Table;Table2.docx ±
Table 2 - Mean D50 (^m) and standard déviation determined by the devices without US. n.a. not available.
BCR66 BCR67 SRM1003C Mix_A Mix_B Mix_CCertified range (^m) 0.35 - 3.5 2.4 - 32 25 - 50 n.a. n.a. n.a.Certified D50 (^m) 1.13 10.41 32.1 ± 1.0 n.a. n.a. n.a.Cilas 1.2 ± 0.1 12.4 ± 0.1 32.6 ± 0.8 11.5 ± 0.1 4.2 ± 0.4 11.4 ± 0.5Coulter 1.5 ± 0.1 11.3 ± 0.1 31.6 ± 0.1 8.9 ± 0.5 1.8 ± 0.1 4.6 ± 2.3Malvern 1.9 ± 0.1 12.3 ± 0.1 31.9 ± 0.1 10.1 ± 0.1 3.0 ± 0.1 8.9 ± 0.1Portable C 1.9 ± 0.1 11.8 ± 0.4 32.6 ± 0.6 9.3 ± 0.1 2.8 ± 0.1 5.9 ± 0.7
Table3 Click here to access/download;Table;Table3.docx ±
Table 3 - Mean proportion (%) and standard déviation of the simple classes measured by the four devices on the natural samples with and without US. n.d. not detected; n.a. not analyzed.
Barcarin Jons Azergues ArcDevices US Clay Silt Sand Clay Silt Sand Clay Silt Sand Clay Silt SandCilas without 8.8 ± 1.3 90.3 ± 0.8 0.5 ± 0.5 3.2 ± 0.4 69.2 ± 0.4 27.4 ± 0.9 2.0 ± 0.1 51.3 ± 0.7 47 ± 0.9 0.7 ± 0.1 1.0 ± 0.1 98.3 ± 0.5Cilas with 9.8 ± 2.1 90.0 ± 1.9 0.2 ± 0.4 4.4 ± 0.5 78.3 ± 1.6 17.1 ± 1.5 3.0 ± 0.1 87.6 ± 0.7 9.4 ± 1 1.0 ± 0.1 6.4 ± 2.0 92.8 ± 1.9Coulter without 21.6 ± 0.9 78.0 ± 0.7 0.4 ± 0.5 8.0 ± 0.1 64.8 ± 1.0 27.3 ± 1.0 2.0 ± 0.1 49.0 ± 1.5 49.4 ± 1.4 0.7 ± 0.5 13.0 ± 8.1 86.3 ± 8.6Coulter with n.a. n.a. n.a. 10.7 ± 1.2 67.3 ± 3.8 21.7 ± 5.0 2.0 ± 0.1 55.3 ± 1.1 42.7 ± 1.1 0.7 ± 0.5 14.2 ± 8.4 84.9 ± 8.8Malvern without n.a. n.a. n.a. n.a. n.a. n.a. 1.7 ± 0.5 44.2 ± 2.1 54 ± 2.2 n.d. n.d. 100.0 ± 0.1Malvern with n.a. n.a. n.a. n.a. n.a. n.a. 2.8 ± 0.4 62.3 ± 3.9 34.9 ± 4.3 n.d. 0.4 ± 0.9 99.6 ± 0.9Portable_ C without 11.0 ± 1.4 87.5 ± 0.7 1.5 ± 2.1 3.0 ± 0.1 74.0 ± 2.8 23.0 ± 2.8 2.1 ± 0.6 72.6 ± 9.9 25.1 ± 10.7 0.3 ± 0.1 9.6 ± 1.0 90.0 ± 1.0Portable C with n.a. n.a. n.a. 7.0 ± 0.1 89.3 ± 2.1 3.3 ± 3.2 4.2 ± 0.4 89.4 ± 1.7 6.4 ± 2.1 0.7 ± 0.8 15.4 ± 11.8 83.9 ± 12.3
Table4 Click here to access/download;Table;Table4.docx ±
Table 4 - Mean D50 (|im) and standard déviation measured by the four devices on the natural samples with and without US. n.a. not analyzed.
Devices US Barcarin Jons Azergues ArcCilas without 9.9 ± 0.5 34.6 ± 1.0 57.0 ± 1.3 194.1 ± 3.6Cilas with 9.2 ± 0.7 25.0 ± 1.3 23.1 ± 1.1 179.3 ± 5.3Coulter without 6.2 ± 0.4 28.2 ± 0.7 61.1 ± 2.8 146.6 ± 20.0Coulter with n.a. 21.9 ± 3.7 49.7 ± 1.3 145.9 ± 20.5Malvern without n.a. n.a. 71.6 ± 5.8 171.9 ± 1.6Malvern with n.a. n.a. 40.0 ± 5.4 171.5 ± 2.1Portable_C without 7.9 ± 1.0 35.4 ± 0.2 37.9 ± 9.1 143.9 ± 20.1Portable_C with n.a. 15.5 ± 3.4 20.9 ± 1.8 132.9 ± 25.3
Figure caption Click here to access/download;Figure;Figure caption.docx ±
Fig.1 Proportion of the detailed classes measured with US by the different devices (Cilas, Coulter,
Malvern and Portable corrected) on the CRMs (BCR-066, BCR-067 and SRM1003c) and the different
mixtures (A, B and C). Certified proportions are also presented (Reference). Values over the bar
represents the percentage of the proportion
Fig.2 Detailed classes measured by the different devices on the natural samples with and without US.
Values over the bar represent the percentage of the proportion. No measures were conducted on the
Malvern for Barcarin and Jons samples
Fig.3 d50 measured by the different devices on the mixtures of Arc and Azergues samples
Fig.4 Detailed classes measured by the different devices on mixture 2 with and without US. Values over
the bar represent the percentage of the proportion
Fig.5 Relative difference between the mean d50 measured by all the devices and the mean d50
measured by each device with and without US. Grey areas represent 10% and 20% difference
Fig.6 Relative differences between the mean proportion of each simple class measured by all the
devices and the mean proportion measured by each device with and without US. Grey areas represent
10% and 20% difference. Abscise represent the mean proportion of the associated simple class for all
devices
Fig.7 Relative differences between the mean proportion of each detailed class measured by all the
devices and the mean proportion measured by each device with and without US. Grey areas represent
10% and 20% différence. The abscissa represents the mean proportion of the associated detailed class
for all devices
FigurelDevices and Reference: Cilas Coulter Z Malvern
Click here te Portable_C
ac cess/downloadReference
Figure;Fig1.pdf ±
100%
50%(/>
0%
100%~o
14 70 8753 58
23 24 37 3313 3 66 7 0 00110 00200 00100 00000
0) 50%
0%
100%0)
50%
4 6 5 5 0 4 6 4 8 316 19 17 20 27 40 39 39 38 —31 27 30 17 4 3 5__5 0 0 0 0 0 0
O(00)
4—O
0%
41 41 47 46 45 49 52 52 38 55
OCMOCO
CO
0 10 10 0 10 10
0"10CO
CM 5 0 1 6 0
___ 14 20 17 19 8 12 12 14 13 14 13 14 _30 26 26 26 28 24 25 22---------!-- 1------ 1------1 1------11^^ 7 4
00
4 0 0 0 0
Mix_A
I--------------- I 1________11111------ 1------1------ 1------1 1§ 100%
50%
0%
100%
û_ 50%
0%
5435 35 39
Z ^■25 22B23 10 ^ 17 ^ 10 ^ 12 20 4 13 14 ^8 15 11 3 0 6 1 0 0 0 0 CD
2134
■ I A 0000 O
Clay V_F_Silt F_Silt M_Silt C_Silt V_C_Silt V_F_Sand
0
Figure2Devices: Cilas Coulter Z Malvern Portable_C
Click here to access/download;Figure;Fig2.pdf ±
100%
50%
0%100%
(/>JS 50%O
-O 0%100%
22 11 12 15 17 18 21 24 30 21 24___19 15 i4 10 5 7001 000 000 000
383 4 5 6 6 9 814 13 12 20 17 19 26 20 22 17 20
5 10 2 0 0 0 0 0 0
(0 50%
0%100%
50%(0
0%'g 100%
50%O
0%100%
Q.50%
û_ 0%
100%
50%
0%
11 7 6 7 12 8 11 16 17 14 18 21 18 23 25 18 _20 16 143 1 71 7 0 0 0 0 0 0 0
2 2 2 2 2 2 2 4 4 4 47 9 8 8 13 l^lil^?1—2liii8r27 22 22 22 19 16 16 17 5 9 8 11 Q
0340
CDCO £= CD
C/)
11 19 11 19 4°16 19^4 4^22 23 24 21 17 ^ 123234 4 3 3 7 85 6 41 10 H 16 9 6 0 ZAZ 0 0760 0230
CDCQ
dCD
(/>
57 63
0100 0101 0101 0201 0302 15050520 22 21 34 4̂ enzz -32
7 16 810 0 0
1101 0101 1 0 2 3 0 A 2 4 0 A 2 5 0 , -T f21 22 20
34 4256 63æ 43
29 7 16 610 0 0
Clay V_F_Silt F_Silt M_Silt C_Silt V_C_Silt V_F_Sand F_Sand M_Sand C_Sand
9
4
Figure3 Click here to access/download;Figure;Fig3.pdf ± US O without Q with
90
60mû
30
Cilas Coulter Malvern Portable CO O O O O
O * O
K>H
HZH
1-
0-1
i-EH
1-
0-1
O
1-0-
1
□ °
**
*
□□□□□ OOoD□□
\ q, \ q, \ q, \ q,
Figure4Pr
opor
tion
of e
ach
deta
illed
clas
sClick
Devices: Cilas Coulter Malvern Portable_Chere to access/download;Figure;Fig4.pdf ±
19
7
2 3 1 1 2 ---- 2 2 ---- 3=c
20
163 143 114ÛJ[
2521
19J
23 22â
24 24
*:l
3L
19—
111
7^83 3 2 4 4 4 3 5
2926
19 20tl
2524 22 22H 19
1310
10 2
___ 5
0
84
20 0 0
Clay V_F_Silt F_Silt M_Silt C_Silt V_C_Silt V_F_Sand F_Sand M_Sand C_Sand
Figu
re5
Clic
k he
re to
Dev
ices
: O C
ilas O
Cou
$fôi
esgÿ
do\M
fe)|¥
èiFr
feui
©;F
iffep
çMbl
e
i i
(%) aouajajjip aApeiay
Mea
n D50
(pm
)
200
100
0
■100
200
100
0
■100
classes: O Clay □ Silt O SandClick here to access/download;Figure;Fig6
Devices: • Cilas • Coulter • Malvern • Portable C
£<>
OO
O □ n
=k>
=■=
Se
0 25 50 75
Proportion (%)
pdfW
ithout sonication W
ith sonication
Rel
ativ
e di
ffére
nce
(%)
Figure7• Clay • M_Silt • V_F_Sand • C_Sand
Class: • V_F_Silt • C_Silt • F_Sand • V_C_SandO F Silt O VC Silt O M Sand
Click here to access/download;Figure;Fig7.pdf| ± Devices: O Cilas □ Coulter O Malvern A Portable C
0
Proportion (%)40
Supplementary Material
Click here to access/downloadSupplementary Material
Electronic Supplementary Material.xlsx
English proofreading
Click here to access/downloadSupplementary Material
proofreading certificate-Grain size analyzers.doc.pdf